BACKGROUND: Within the context of a European network dedicated to the study of sarcoma the relevant literature on sarcoma risk factors was collected by searching PubMed and Google Scholar, the two information storage and retrieval databases which can be accessed without charge. The present study aims to appraise the relative proficiency of PubMed and Google Scholar. FINDINGS: Unlike PubMed, Google Scholar does not allow a choice between "Human" and "Animal" studies, nor between "Classical" and other types of studies. As a result, searches with Google Scholar produced high numbers of citations that have to be filtered. Google Scholar resulted in a higher sensitivity (proportion of relevant articles, meeting the search criteria), while PubMed in a higher specificity (proportion of lower quality articles not meeting the criteria, that are not retrieved). Concordance between Google Scholar and PubMed was as low as 8%. CONCLUSIONS: This study focused just on one topic. Although further studies are warranted, PM and GS appear to be complementary and their integration could greatly improve the search of references in medical research.